Intercontinental Diversity of Caballeronia Gut Symbionts in the Conifer Pest Bug Leptoglossus occidentalis

Abstract

Many stinkbugs in the superfamily Coreoidea (Hemiptera: Heteroptera) develop crypts in the posterior midgut, harboring Caballeronia (Burkholderia) symbionts. These symbionts form a monophyletic group in Burkholderia sensu lato, called the "stinkbug-associated beneficial and environmental (SBE)" group, recently reclassified as the new genus Caballeronia. SBE symbionts are separated into the subclades SBE-α and SBE-β. Previous studies suggested a regional effect on the symbiont infection pattern; Japanese and American bug species are more likely to be associated with SBE-α, while European bug species are almost exclusively associated with SBE-β. However, since only a few insect species have been investigated, it remains unclear whether region-specific infection is general. We herein investigated Caballeronia gut symbionts in diverse Japanese, European, and North American populations of a cosmopolitan species, the Western conifer seed bug Leptoglossus occidentalis (Coreoidea: Coreidae). A mole-cular phylogenetic ana-lysis of the 16S rRNA gene demonstrated that SBE-β was the most dominant in all populations. Notably, SBE-α was rarely detected in any region, while a third clade, the "Coreoidea clade" occupied one fourth of the tested populations. Although aposymbiotic bugs showed high mortality, SBE-α- and SBE-β-inoculated insects both showed high survival rates; however, a competition assay demonstrated that SBE-β outcompeted SBE-α in the midgut crypts of L. occidentalis. These results strongly suggest that symbiont specificity in the Leptoglossus-Caballeronia symbiotic association is influenced by the host rather than geography, while the geographic distribution of symbionts may be more important in other bugs.

Keywords: Caballeronia; intercontinental diversity; obligate gut symbiosis; stinkbug.

Fig. 1.

The conifer bug Leptoglossus occidentalis and its midgut structure. (A) An adult of L. occidentalis on the leaves of a pine tree and (B) its whole midgut structure. The bug shown was reared in the laboratory. Symbiont inoculation was performed by feeding the insect a soil suspension. A symbiont native in the soil colonizes the midgut crypts (M4). Abbreviations of the midgut sections are as follows: M1, midgut first section; M2, midgut second section; M3, midgut third section; M4, midgut fourth section with crypts; M4B, M4 bulb; H, hindgut.

Fig. 2.

Molecular phylogenetic ana­lysis of Caballeronia gut symbionts of the conifer bug Leptoglossus occidentalis. A maximum-likelihood tree was generated based on 1,256 aligned nucleotide sites of the 16S rRNA gene. Numbers at the tree nodes indicate the maximum-likelihood bootstrap values (%) with 1,000 replicates, and bootstrap values of more than 50 are shown. We referred to the nucleotide sequence information reported in previous studies on the Caballeronia gut symbionts of coreoid insects in Japan (Kikuchi et al., 2011a; Kuechler et al., 2016; Ohbayashi et al., 2019b), in America (Olivier-Espejel et al., 2011; Garcia et al., 2014; Acevedo et al., 2021; Hunter et al., 2022), and in Europe (Kuechler et al., 2016; Ohbayashi et al., 2019b). The subtree of the SBE-α group is compressed. An uncompressed subtree is shown in Fig. S1. Accession numbers in the DNA database (DDBJ/EMBL/GenBank) are shown in square brackets. L. occidentalis gut symbionts are shown in blue with bold case font. Stars: bacterial strains used for symbiont inoculation tests. GS: Gut symbiont.

Fig. 3.

Relative abundance of SBE-α, SBE-β, and Coreoidea clade bacteria among gut symbionts of conifer bugs normalized by one OTU by an individual at the country level. The number of investigated insects in each country is shown in the graphs, and the precise numbers are provided in Table S1 and S2. Relative abundance at the local level is shown in Fig. S2.

Fig. 4.

Leptoglossus occidentalis infection with SBE-α and SBE-β strains and their host fitness effects. (A, B, and C) Whole midguts of L. occidentalis at the 3^rd instar stage, and (D, E, and F) an enlarged image of midgut crypts. A dissected midgut inoculated with (A and D) an SBE-α GFP strain, with (B and E) an SBE-β GFP strain, and without (C and F) any inoculant (aposymbiotic). An abbreviation of the midgut section is as shown in Fig. 1. (G) Survival rates of L. occidentalis inoculated with SBE-α (blue line, n=13) and SBE-β (green, n=20), and without any inoculant (aposymbiotic: gray, n=27). The survival period was followed from hatching to the last adult emergence. A black arrow and dotted line indicate symbiont infections at 6 days post-hatching. Different letters indicate significant differences (P<0.0001, Fisher’s exact test with the Bonferroni correction). (H) Developmental time from hatching to adulthood in conifer bugs inoculated with SBE-α (blue bar, n=13) and SBE-β (green, n=18), and without any inoculant (aposymbiotic: gray line, n=1, i.e. the only surviving insect in the experiment). The mean±SD is shown. n.s. indicates no significant difference (Student’s t-test).

Fig. 5.

Competition assay of RFP-labeled SBE-α and GFP-labeled SBE-β strains in midgut crypts of Leptoglossus occidentalis. (A) The midgut crypts of L. occidentalis at 7 days post-infection (dpi), infected with an equal mixture of both strains. A merged GFP and‍ ‍RFP image is shown. (B) Relative abundance, determined by flow cytometry, of GFP and RFP strains at the inoculum and at the midgut crypts at 7 dpi (n=10). The abundance of GFP and RFP strains at the midgut crypts is significantly different (P<1×10^–10, the Student’s t-test). Fluorescent microscopic images and the relative abundance of GFP and RFP strains in 10 individual insects are shown in Fig. S3. Note that both strains resulted in 100% infection when used in control mono-infections (Fig. S4).